Almost 12 noon and the average temp in the USA is under 16 degrees..can you imagine that? OK...As far as temps go..brutal overnight again. 20s tomorrow 30s thursday with perhaps light snow overnight 30s/40s Friday with chance of light snow 40s/50s Saturday with rain 50s Sunday Maybe 60 Monday with a chance of rain After that, cools off but nothing like this..maybe middle to end of the week next week snow even.
Correct - like any average it is the integral of the value of the random variable times the density function of the random variable over the domain of interest. I expect in this case it's a simple unweighted mean of the temperatures over all of the official weather stations in the continental US, which is more than accurate enough as long as the weather stations are spread over the US reasonably uniformly.
and thankfully Jake...as we speak..that little bitch known as the polar vortex has heard your call and is back on its way to you...RELIEF DOWN HERE. In fact..if you can believe this..temps will actually RISE for many overnight..now rise might be its 3 now and it rises to 6 before sunrise. Anyway, we start to thaw out tomorrow..20s...30s...40s...50s...just put the next day to it. Now, I have been lack checking the models really well just glancing. When I se there are no storms quite honestly, I get a bit bored. So, I just did a quick check and really need to study what the hell is going on because here are the options. Saturday rain..maybe heavy...50s. Sunday night/Monday a noreaster..but no real cold air? I mean some models have a real bomb..some have it going out to sea. Then another storm Wednesday next week..and another one Saturday next week. The only way I can really read it right now is the chance that any of these are snow increase as the time goes. Saturday is rain...Monday should be rain...Wednesday could be rain...could be snow...and then Saturday. Very poor breakdown and tonight I will have a write up..promise.
Bingo Staff..and I completely agree with Cman..I read the first sentence and while I began to drool during it...I was all in sir.
first off models suggest a possible light snow event. Problem is the EURO brings it too weak and too far south so its a low percentage chance for even light snow..that would be Thursday night. Then the Euro and GFS agree in a storm coming up to our west..warmer storm on Saturday which would bring us rain..and perhaps an inch of it. Did not see the Euro take on precip amount but the GFS seemed pretty consistent. Then a possible storm on Sunday night/Monday..but there is NO agreement here. Third of the models have no storm..third have warm storm...and the rest have everything you can imagine. Heres the bottom line. Unless something pops and makes sense, when your models do this cha cha, 80 percent of the time its just a simple rain event. The only thing I just saw that is interesting it the pattern we were in that brought us all of the snow looks to reload by around the 20th. I think I posted that several days ago about the return to that pattern..and that pattern will return in around 2 weeks. So lets see if something surprises us before..wouldnt be shocked.
This is assuming that the weather stations are indeed distributed uniformly which I suspect is hardly the case. I would think if the inputs were weather stations that you would need to weight them according to a formula which takes into account proximity to the next nearest station and/or density of weather stations in a fixed area such as a 100 mile radius
I imagine that weather stations are less densely packed west of the Mississippi, but the biggest problem with that would be if there are isolated areas between stations with very different temperatures than the ones at the closest stations, which seems unlikely to me (if that was the case that means the weather patterns are noticeably different, and as a result the NWS would put a weather station there). A bigger problem would be if there are lots of stations in big cities, since cities are going to be warmer than the surrounding suburbs because of the heat reflected off the pavement. I would hope that they would take that into account somehow (or have confirmed that it doesn't really matter). Jay might know if they put lots of stations in big cities or not. Maybe they don't - after all, in New York it's always the temperature in Central Park or at JFK that they focus on. Anyway, as you say, a weighted version that uses the geographic density of stations would be more accurate.
This is the "smart" way to go: http://en.wikipedia.org/wiki/Satellite_temperature_measurements But I'm guessing the number quoted in popular media is probably a "dumb" average
While you are dead on staff...I would also think many are in big cities..but it would be less effective in a normal pattern to average the mean temp or a spring or fall pattern because lets say the average temp in the country was 57 in April...you might have temps in the midwest or northern new england in the 20s compared to Florida, Texas, Nevada etc. that might be in the 80s..so the average means little because its not telling any story other than south is warm and north is colder. This map..for this particular basically country wide freeze is rare and good accurate time to see it encompasses pretty much the entire country. Where they are located and if it was evenly distributed doesnt mean nearly as much. Perfect example is this: Here, whether the stations are in lets say Green Bay...or bumf#$#, WI. or the stations are evenly distributed down south lets say, with this particular freeze, I would think it would make little difference. Looking at this map to me, it seems like almost 90% of the country is below freezing right now. I think you guys are dead on though in a normal pattern. The average mean in the country with the way they do this is pretty meaningless unless the country is either in a deep freeze..or a heat wave for the most part. Not sure if that makes sense or if Im explaining that well because I just got home after a half bottle of wine. One other thing..you mention the cities heating staff..another excellent point. But again, with this type of freeze, other than their actual location, the cities have not benefited from their concrete environment etc with this freeze. Temps in the cities have been pretty consistent with their suburban neighbors for the most part which just goes to show how extreme this cold spell has been.
Did a quick check on models..and not much of a change. The models keep hinting at above average temps next week which would make anything coming through a rain event...but thats before we get through Thursday and Friday. Just saw the Canadian model which gives a rolling future radar and it continues to hint at two things. A light snow event for Friday that might bring a quick 1 to 3 or 2 to 4 event..and the rain event for Saturday starting out as frozen precip. Unlike that last storm where the cold air wouldnt budge and we ended up having a prolonged freezing rain event last weekend, that should not be the case because this current airmass will be will back on its way to Canada by Thursday. Here is the model I just spoke about...good stuff. You can hit the ANIMATE button and watch it roll out hour by hour...goes out about 120 hours from right now. You will first see the blue come through for Friday...followed by a quick red (freezing rain) for Saturday that is erased quickly with a lot of green (rain) for Saturday night/Sunday. After this storm goes through, you will see not much..but a progressive pattern that should allow temps to rebound next week to a bit above normal for at least part of the week. Think we might be around 14 days away from getting back into a snowy pattern. LINK: AGAIN..JUST HIT THE ANIMATE BUTTON: http://collaboration.cmc.ec.gc.ca/cmc/cmdn/pcpn_type/pcpn_type_gem_reg.html
Yes, I agree completely. Usually this is the classic joke "If you encounter a statistician with one foot in a bucket of ice and the other in a bucket of boiling water, he will say that on average he feels fine." It just so happens that in this case it was pretty much the entire country that was affected by the cold, so the number was meaningful.
:chargers: dead on staff...dead on. guys..have some personal issues with mom..real sorry I have not been around. Beware of frozen precip tonight and just saw major rains for Saturday with flooding. Will try to get on tonight and again sorry.
